PORQUE ESTA EN LA ARCADIA

Mean value of Size (in terms of total assets) of the banking industry is $8 billion, $13 billion and 10 billion in Pakistan, China and India respectively which shows that banking industry in Pakistan is not as large as in China and India (Table II) . It is fluctuated between $6 to $9 billion in Pakistan $10 to $14 in China and $8 to $11 in India. The statistics of

Table 5.1: Descriptive Analysis

Mean Standard Deviation Minimum Maximum

Variable Pakistan India China Pakistan India China Pakistan India China Pakistan India China

Size 8.5811 10.417 13.063 .62761 .73766 1.0978 6.9006 8.342 10.204 9.6882 11.44 14.929 CRK .04888 .01682 .00778 .01294 .03225 .0131 .01535 0 .0008 .07537 .175 .1247 CR8 .41120 .21136 .45247 .03705 .02059 .07832 .36054 .178 0.354 .47878 .24 0.57 ICRG 6.6181 8.644 7.2412 1.393 .3315 .49159 4 8.167 6.5 8 9.333 7.988 LY 6.7944 6.9447 8.069 .25011 .35038 .56848 6.3026 6.337 7.149 7.1708 7.371 8.8607 EBTSE .02738 .02340 .01587 .01214 .00720 .00417 .00357 .006 .0046 .06172 .038 .02270 EBTSLP .08306 .02606 .01686 .02234 .00653 .00795 .0241 .01602 .0054 .14021 .05765 .08242 NIM .08386 .03704 .03054 .03088 .04727 .00932 .0342 .002 .0019 .14794 .381 .06654 Source: Author’s calculation

Pakistani banking industry for credit risk is fluctuated between .015 and .075 and Mean value is .048. For China, same variable is fluctuated between .0008 and .1247 while in India it varied from 0 to .175. Results for market concentration reveals that mean value stands at .411, .452 and .211 in Pakistan, China and India respectively.

Mean value of Governance 6.61, 7.24 and 8.26 respectively in Pakistan. China and India. It is fluctuated between 4 to 8 in Pakistan, 7 to 8 in China and 8 to 9 in India. The average value of GDP is 6.7 in Pakistan 8.06 in China and 6.94 in India.

5.2 Empirical Analysis

Profitability model explains that not only bank’s related variables affects the banking profitability but also some other variables too which can affect the profitability. For analysis we have used SGMM and DGMM. In the Asian region three economies with emerging banking markets have responded to these variables. SGMM analyzes that size of the banking industry has significant relation with respect to bank profitability especially in case of Pakistan because Pakistan’s banking industry is little bit behind as compared to other economies with emerging banking markets.

Table 5.2a: Regression Estimation For Profitability Dependent Variable: NIM ,s

Pakistan India China

Variable DGMM SGMM DGMM SGMM DGMM SGMM NIM 0.223 (0.078) 0.826 (0.000) 0.187 (0.000) 0.126 (0.000) 0.318 (0.000) 0.475 (0.000) S -0.034 (0.000) 0.006 (0.002) -0.009 (0.620) -0.012 (0.309) -0.013 (0.001) -0.000 (0.723) CRK 0.204 (0.135) -0.091 (0.674) -0.203 (0.003) -0.160 (0.074) 0.526 (0.000) 0.446 (0.001) CR8 0.114 (0.066) -0.056 (0.442) 0.463 (0.368) 0.460 (0.367) 0.018 (0.441) -0.036 (0.035) GOV 0.011 (0.000) 0.002 (0.047) 0.014 (0.006) 0.020 (0.079) -0.000 (0.758) 0.003 (0.001) LY 0.015 (0.229) -0.032 (0.026) 0.054 (0.240) 0.070 (0.221) 0.016 (0.004) 0.010 (0.001) AR(1) 0.892 0.028 0.160 0.112 0.021 0.014 AR(2) 0.592 0.134 0.303 0.306 0.383 0.261 Hansen test 0.855 0.948 1.000 1.000 1.000 1.000

Source: Authors’ calculations using stata (12.0) command Xtabond, robust

In Pakistan size of the banking industry is directly related to the profitability. These results follow the economies of scale in Pakistan’s banking industry (Gilchrist, 2012). According to the results size measured by total assets shows that as size of the industry keeps on growing, banks have more opportunities to invest and thus the deposit rates goes down as compared to lending rates and in this way profitability of banks increase, this variable is significant at 1% significance level. It means that size of the banking industry has better capacity to explain variation in profitability of banking sector in Pakistan. But as compared to Pakistan, China and India both the countries have larger banking industries and have showed no significant relationship between size and profitability. Heffernan and Fu (2008) found no relationship between size and profitability because up to a certain limit, size has affected profitability. After this breakeven point, it loses its impact. So our results match with their findings in case of China and India. It means that in India and China banking industries have already reached to its maximum limit therefore with the increase in size profitability will not increase.Our DGMM results are in line with SGMM. So our first hypothesis is not rejected in case of Pakistan which states that there is a link between size and profitability of banking industry.

Another important factor for bank profitability is credit risk which is measured by loan loss provisions to total advances used by many authors such as (Athanasoglou et al., 2008) and (Tan & Floros, 2012). In Pakistan, credit risk has insignificant negative impact on the profitability of banking industry. Both our techniques have shown similar results and these results support the findings of (Sayedi, 2014). But in China and India, credit risk has significant impact on bank profitability. For China, credit risk has positive significant results with profitability and significant at 1% significance level according to the both of GMM techniques and supports the findings

of(Gizaw et al., 2015). Since banks have already provided for the losses therefore recovery tactics are used for NPLs which directly hit the profitability of banks. In contrast, Indian banking sector has showed negative response towards credit risk and showed it is significant at 10%. These results support the findings of (Gunter et al., 2013) that an increase in loan loss provisions decrease the profitability of banking sector.

Concentration has no strong impact on the profitability of banking sector in respective economies except China. In China concentration has negative significant relationship with profitability of banking sector and supports our hypothesis stated that there is a relationship between concentration and profitability. García-Herrero et al. (2009), also found negative sign between concentration and profitability in Chinese banking sector and argued that high concentration leads to increase in NPLs due to imprudent lending practices which ultimately declines the profitability of banking industry. Concentration is statistical significant under DGMM in Pakistan and support the findings of(Bourke, 1989) but these results are not significant.

ICRG (International Country Risk Guide) is combination of three elements but here one of the components is used as a measure for Governance is investment profile. The standard of this index tells that as rating of the component is high risk will be low and vice versa. So when investment profile has more rating it means the country has enjoying favorable conditions and more foreigners will show interest for investment (Hayakawa, Kimura, & Lee, 2013). According to the results ICRG has positive influence on bank profitability. Both SGMM and DGMM showed its significance for the bank profitability in all the three countries. And these results are significant at 10%, 5% and 1% significant level. These results supported our hypothesis which states that there is a relationship between governance and bank profitability.

Macroeconomic conditions also affect bank profitability such as GDP. GDP has a mixture of significant impact on the respective countries except India under GMM techniques. In Pakistan GDP has negative impact on bank profitability and the results are significant at 5%. The results supports the findings of (Bonin, Hasan, & Wachtel, 2005; Demirgüç-Kunt & Huizinga, 1999), operating cost decreases as GDP increases so this will lead to narrow the interest margins thus negative relation exist between NIM and GDP (Azeez & Gamage, 2013). But In China GDP has positive significant impact on profitability of banks. These results show support for the findings of (Ali et al., 2011; Goddard et al., 2004).Positive sign indicates that when GDP increases economy grow and projects required more financing from banks. This variable is significant at 1%. DGMM also shows same results in China. These results supports the hypothesis regarding GDP.

Table 5.2b: Regression Estimation For Expense Preference Theory Dependent Variable: EBTSE ,s

Pakistan India China

Variables DGMM SGMM DGMM SGMM DGMM SGMM EBTSE,t-1 -0.049 (0.752) 0.592 (0.000) -0.584 (0.205) 0.985 (0.000) 0.522 (0.001) 0.310 (0.044) S -0.009 (0.088) 0.005 (0.000) -0.001 (0.624) -0.002 (0.162) -0.006 (0.001) -0.000 (0.934) CRK -0.268 (0.000) -0.283 (0.004) -0.103 (0.025) -0.038 (0.054) 0.087 (0.184) 0.063 (0.043) CR8 0.095 (0.033) 0.075 (0.083) -0.014 (0.436) 0.009 (0.640) 0.017 (0.001) 0.001 (0.904) GOV 0.007 (0.000) 0.003 (0.000) 0.004 (0.509) -0.008 (0.004) -0.001 (0.042) 0.000 (0.799) LY -0.005 (0.385) -0.088 (0.271) 0.003 (0.694) -0.004 (0.094) 0.006 (0.029) 0.002 (0.210) AR(1) 0.106 0.032 0.209 0.036 0.044 0.029 AR(2) 0.057 0.196 0.578 0.258 0.247 0.007 Hansen test 0.813 0.993 0.391 0.998 1.000 1.000 Source:Author’s calculations using stata (12) command Xtabond, robust

P values are in parenthesis

Value added measures are being used in this study to analyze the managerial behavior towards expense preference as well as risk avoidance. For this purpose earning before tax and staff expense to total assets is used for expense preference behavior

and earning before tax, staff expense and loan loss provisions to total assets used for risk avoidance behavior as used earlier by (Bourke, 1989). In Pakistani banking industry relationship between concentration and dependent variable (EBTSE) is significantly positive which shows support for the existence of expense preference theory. Both SGMM and DGMM shows similar results and significant at 10% and 5% significance level respectively, and these results are in contradictory with previous studies such as (Bourke, 1989). In India results show no supportive evidence for this theory. In china under DGMM results show support for expense preference theory and significant at 1% but under SGMM results did not show strong relationship between concentration and EBSTE. In Pakistani banking industry, size has positive impact on the dependent variable (EBSTE) and significant at 1%. This indicates that when size of the banking industry increases it would lead to increases in staff expenses, but in China and India it has insignificant relation. Credit risk also has significant relationship in all the three countries under SGMM. In Pakistan and India credit risk has negative correspondence but in China it has positive significant relation. Governance have its impact on bank’s staff expense, in Pakistan it has positive impact showing that as country enjoying better environment for investment it ultimately increases profitability of banks and staff salaries too. But opposite in India governance has negative significant impact which shows that as country facing less favorable conditions or highly risky environment for investment, bank management have to face a challenge of staff maintenance. To retain their experienced staff with their self they increase salaries. GDP has no any significant relationship in any of the respected country at 5% and 1% significance level.

Dependent Variable: EBTSLP

Pakistan India China

Variables DGMM SGMM DGMM SGMM DGMM SGMM EBTSLPi,t-1 -0.053 (0.847) 0.587 (0.012) 0.223 (0.034) 0.393 (0.001) -0.088 (0.322) 0.113 (0.049) S -0.023 (0.005) 0.008 (0.043) 0.006 (0.257) 0.001 (0.504) -0.001 (0.192) -0.000 (0.517) CRK -0.158 (0.235) -0.434 (0.002) 0.091 (0.366) 0.026 (0.438) 0.526 (0.000) 0.428 (0.000) CR8 0.366 (0.000) 0.267 (0.000) -0.032 (0.308) -0.040 (0.273) 0.025 (0.005) 0.009 (0.308) ICRG 0.015 (0.000) 0.007 (0.006) -0.000 (0.943) -0.001 (0.498) -0.000 (0.260) 0.000 (0.688) LY 0.010 (0.431) -0.005 (0.667) -0.012 (0.210) -0.004 (0.256) 0.001 (0.505) 0.003 (0.152) AR(1) 0.304 0.031 0.121 0.166 0.041 0.048 AR(2) 0.655 0.101 0.406 0.359 0.881 0.419 Hansen test 0.737 0.990 1.000 1.000 0.894 1.000 Source:Author’s calculations using stata (12) command Xtabond, robust

P values are in parenthesis

Another value added measure which is being used to test (EHM Hypothesis) is EBSTLP. In our third model EBSTLP is dependent variable. The results show that in Pakistan and China although concentration is being significantly affects the dependent variable under SGMM and DGMM but signs indicate that there is positive relationship. Not even in SGMM but also under DGMM technique in Pakistan. Which is not in line with the findings of (Bourke, 1989) but in India although the results are insignificant under SGMM and DGMM but negative sign shows support for Edward Heggestad Mingo Hypothesis. It means that when concentration increases it would lead to decrease the loan losses which show that management’s behavior is purely risk averse. But in Pakistan and China results show opposite signs and indicate that there might be more dependence of managers on stockholders of the company that is why in more concentrated banking industry managers might be less risk averse. In Pakistan Size also matters in this case and affects it positively because when Banks increase their size, it ultimately increases its number of employees and staff expense also increases but in China and India it has insignificant impact. Credit risk in China has positive impact because when Credit risk increases it would

lead to increase in loan loss provisions so there is positive relation between these two.

Table5.2d: Hypothesis Testing

Hypothesis Pakistan China India

There is link between size and profitability Supported Rejected Rejected There is relationship between Credit risk and profitability Rejected Supported Supported There is a link between concentration and profitability Rejected Rejected Supported There is a connection between GDP and profitability Supported Supported Supported There is a relationship between Governance and profitability Supported Supported Supported Banking industries of different countries have shown same

behavior under concentrated market

Pakistan

Rejected China

India Source: Authors’ formulation

CHAPTER 6

CONCLUSIONS, RECOMMENDATIONS AND FUTURE